I will get back to the moving average representations and other analysis soon but I wanted to try and think about this from a different visualization. In general I think it is accepted with this, and any disease for that matter, that you get exposure to the disease which shows up with positives, positives lead to hospitalizations, and hospitalizations lead (unfortunately) to deaths. So how are the county positives looking? I look at two different measures, the seven day total positivity rate, and the week-over-week growth in positives.

The color coding represents the 7 day positivity level. Higher levels on each of the scales is a bad thing. We can see there are two outlier counties right off the bat. There is a significant amount of congestion in the 15% level of the positivity rate, which is not good. The week over week growth can be positive or negative and can obviously fluctuate wildly based on positive levels and testing. The congestion there due to the outliers makes interpretation a bit difficult overall, so I remove Foster and Golden Valley.

Removing the two outliers does decompress the data. Some counties are positive in both metrics, like Grand Forks and Williams, while Ward is negative in the week over week positives while still very high on the seven day positivity rate. These two metrics could be considered a starting point for a larger situational analysis determining the direction of later measures like hospitalizations and deaths.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.